An Ultra-Wideband Sensing Board for Radio Frequency Front-End in Iot Transmitters

electronics

Article An Ultra-Wideband Sensing Board for Radio Frequency Front-End in IoT Transmitters

Alessandra Petrocchi 1,2, Antonio Raﬀo 2,*, Gianni Bosi 2 , Gustavo Avolio 3, Davide Resca 4, Giorgio Vannini 2 and Dominique Schreurs 1

1 Department of Electrical Engineering (ESAT), KU Leuven, 3001 Leuven, Belgium; [email protected] (A.P.); [email protected] (D.S.) 2 Department of Engineering, University of Ferrara, 44122 Ferrara, Italy; [email protected] (G.B.); [email protected] (G.V.) 3 Anteverta-mw B.V., 5656AE Eindhoven, The Netherlands; [email protected] 4 Microwave Electronics for Communications, MEC S.r.l, 40127 Bologna, Italy; [email protected] * Correspondence: antonio.raﬀ[email protected]

Received: 9 September 2019; Accepted: 15 October 2019; Published: 19 October 2019

Abstract: The upcoming technologies related to Internet of Things will be characterized by challenging requirements oriented toward the most efficient exploitation of the energy in electronic systems. The use of wireless communications in these devices makes this aspect particularly important, since the performance of radio transceivers is strongly dependent on the environmental conditions affecting the antenna electrical characteristics. The use of circuits capable of adapting themselves to the actual state of the environment can be a valuable solution, provided that the implemented sensing features have negligible impact on the overall performance and cost of the system. In this work, we present the design and verification of an innovative ultra-wideband sensing board to detect real-time variations of the antenna impedance in transmitters oriented to Internet of Things applications. The proposed sensing board was widely validated by means of small- and large-signal measurements carried out at microwave frequencies.

Keywords: six-port junction; ultra-wideband; Internet of Things

1. Introduction Internet of Things (IoT) has become more and more common nowadays in the research community. Starting from [1,2], where the main concepts of IoT were addressed, the research around this topic has grown and, consequently, additional details have been defined about the requirements (e.g., low cost, energy efficiency, broad bandwidth). The basic idea about IoT is that things or objects become smart, which means that they obtain the ability to interact with each other and work together, even exploiting the capabilities of the internet network. The innovation regarding IoT is not necessarily related to the development of new electronic systems but also to the need to adapt existing systems by adding these smart functionalities. IoT devices will be used in applications for our daily life at home or in the industry. For instance, home devices include smart TVs, speakers, wearables, and smart appliances, whereas technologies dedicated to security systems and to sensor and monitor traffic or weather are dedicated to the industry [3]. The communication between devices can be delegated to common standards, such as multiband WLAN [4,5], although ultra-wideband (UWB) transmissions are nowadays spreading out as an effective solution, especially for short-range communications [6–9]. Their use is defined by strict regulations, since they exploit bandwidths already used by other services and applications and, therefore, the transmitted power must be limited to avoid any interference. For example, the US and

Electronics 2019, 8, 1191; doi:10.3390/electronics8101191 www.mdpi.com/journal/electronics Electronics 2019, 8, 1191 2 of 16

European regulations allow UWB transmissions from 3.1 to 10.6 GHz with a maximum power of 41 dBm/Hz over a bandwidth of at least 500 MHz or 20% of the center frequency [9,10]. − Due to the vastness of possible applications, IoT capabilities should be integrated in low-cost objects and deployed under a huge variety of conditions. An important goal will be to equip these devices with the capability to adapt themselves according to the actual operating condition, in order to provide stable performance regardless of any environmental variation. Considering the radio communication environment for the aforementioned applications, the antenna near-field may change drastically in real operating conditions, and this will affect the antenna input impedance. This variation has the strongest impact at the transmitter side (composed of an antenna, a radio frequency (RF) front-end, and digital signal processing), especially on the RF front-end (e.g., power amplifier (PA)). Indeed, the antenna load variation jeopardizes the PA performance, for example, through reducing efficiency, gain, and linearity. In addition, a strong mismatch condition may have a considerable and detrimental impact on the system reliability, since it could lead to the active devices in the PA being close to their maximum rating specifications [11,12]. This is not the first time that this problem has been faced. Indeed, several approaches have been proposed in the literature to solve the mismatch problem between the PA and the antenna, including enhancement of the circuit architecture [13], reconfigurable matching networks [14], and signal adaptation. These approaches are not all compatible with IoT requirements: for example, making the circuit design more robust by means of an isolator would create additional losses which are not in line with the energy-efficiency goal of IoT. In this work, we will assume the PA is an off-the-shelf component, having no tight specifications in terms of robustness to load variations. We also assume that the load presented to the PA can be adjusted in real time by means of a reconfigurable matching network (RMN) (Figure1). In this scenario, the possibility of monitoring the RF front-end load condition to possibly compensate for load variations by means of the RMN is a crucial aspect for maximizing the overall performance of the system. RMNs can be implemented by using different technologies, such as switches, to select between stubs [15] or capacitors [16], microelectromechanical systems (MEMS) [17], and semiconductor varactors [18,19]. Whatever the choice, sensing of the antenna impedance is required in order to properly tune the RMN. To achieve this, multiple techniques have been proposed in the literature. The most common solutions can be divided into two main groups: those focusing on monitoring the amplitude only and with narrowband specification, and those with phase information and broadband implementation [20,21]. However, both approaches are not compatible with IoT devices. The first is not suitable because of the need for vectorial information over a large bandwidth. The second, despite being broadband, adopts a quadrature down converter which creates additional losses and power consumption. Another example is a commercial gain/phase detector [18,22] to acquire the reflection coefficient of the antenna. Being an active device, it requires a DC supply, which is clearly not suitable for low-cost high-efficiency applications. In this paper, the proposed approach consists of an ultra-wideband passive network able to detect the load variations in terms of both amplitude and phase. It is based on a six-port junction [23,24], which is a passive structure. It has been designed to achieve a bandwidth of 1 GHz, i.e., from 5 to 6 GHz, covering the unlicensed national infrastructure (U-NII) radio bands, the 5 GHz Wi-Fi band, and to be suitable for 802.11ac Wi-Fi access points (ACs). The availability of such a large bandwidth is a great advantage in terms of the possible applications, since it can be used to accommodate different services using the same RF front-end and antenna system, e.g., UWB localization [25] with 5-GHz WLAN capabilities for data transmission. To our best knowledge, this is the first time that such a kind of device is presented for an IoT wideband RF transmitter application. This paper is organized as follows. In Section2, the six-port junction and the detector topology and design are explained as well as the fabrication procedure. In Section3, the sensing board performance is validated by means of measurements. The fabricated device will be tested not only to verify Electronics 2019, 8, 1191 3 of 16 itsElectronics capability 2019, 8 to, x FOR achieve PEER theREVIEW correct impedance information but also in terms of nonlinearity3 of and 16 bandwidth. Finally, in Section4, some conclusions will be drawn. Electronics 2019, 8, x FOR PEER REVIEW 3 of 16

Figure 1. Simplified block scheme of a transmitter.

2. Sensing Board Design Figure 1. SimplifiedSimpliﬁed block scheme of a transmitter.

2. Sensing The block Board scheme Design of the proposed sensing technique for the detection of the front-end load impedance variations is shown in Figure 2. The six-port junction [23,24] solution has been chosen The block scheme of the proposed sensing techniquetechnique for the detection of the front-end load because it is simple to construct, low-cost, meets broadband requirements, and does not require any impedance variations is shown in FigureFigure2 2.. TheThe six-portsix-port junctionjunction [[23,24]23,24] solution has been chosen power supply. All of these aspects are compatible with IoT devices. The concept behind this passive because it is simple to construct, low-cost, meetsmeets broadband requirements, and does not require any structure is well-known in the literature [23,24], and it has already been used in various microwave power supply. All of these aspectsaspects areare compatiblecompatible withwith IoTIoT devices.devices. The concept behind this passive and wireless applications, including reflectometers and direct conversion receivers for structure is well-known in the literature [[23,24],23,24], and it has alreadyalready beenbeen usedused inin variousvarious microwavemicrowave communication. and wirelesswireless applications, applications, including including reflectometers reflectometers and direct and conversion direct receiversconversion for communication.receivers for The power detectors associated with the six-port junction are designed to get the baseband communication.The power detectors associated with the six-port junction are designed to get the baseband voltages which are used to calculate the reflection coefficient of the antenna. They also respect the voltagesThe whichpower aredetectors used to associated calculate thewith reflection the six-port coeffi junctioncient of theare antenna.designed Theyto get also the respect baseband the IoT requirements, i.e., wideband operation related to the adopted bandwidth and no need for a power IoTvoltages requirements, which are i.e.,used wideband to calculate operation the reflection related coefficient to the adopted of the bandwidthantenna. They and also no respect need for the a supply. powerIoT requirements, supply. i.e., wideband operation related to the adopted bandwidth and no need for a power This design covers a bandwidth of 1 GHz, from 5 to 6 GHz. In the following subsections, we will supply.This design covers a bandwidth of 1 GHz, from 5 to 6 GHz. In the following subsections, we will describe the design of these elements in detail. describeThis the design design covers of these a bandwidth elements of in 1 detail. GHz, from 5 to 6 GHz. In the following subsections, we will describe the design of these elements in detail.

Figure 2. Block scheme of the sensing board. Figure 2. Block scheme of the sensing board. 2.1. Six-Port Junction Design 2.1. Six-Port Junction Design Figure 2. Block scheme of the sensing board. The six-port junction can be seen as a black box with one input and five outputs (see Figure3). Four2.1. Six-PortThe output six-port Junction ports junction will Design be connectedcan be seen to as four a black power bo detectorsx with one providing input and four five voltages outputs from (see whichFigure the 3). antennaFour output reflection ports coewillffi becient connected can be retrieved to four power in terms detectors of amplitude providing and phasefour voltages through from an appropriate which the The six-port junction can be seen as a black box with one input and five outputs (see Figure 3). calibrationantenna reflection algorithm. coefficient We will describecan be thisretrieved aspect inin Sectionterms 2.3of .amplitude and phase through an Four output ports will be connected to four power detectors providing four voltages from which the appropriateIn the literature, calibration it isalgorithm. possible to We find will several describe solutions this aspect for the in designSection of 2.3. a six-port junction. In our antenna reflection coefficient can be retrieved in terms of amplitude and phase through an case,In considering the literature, thedemand it is possible of a simpleto find structure,several solutions we adopted for the the design one described of a six-port in [26 junction.]. As shown In our in appropriate calibration algorithm. We will describe this aspect in Section 2.3. Figurecase, considering3, it has one the input demand port, whichof a simple is connected structure, to we the adopted power amplifier the one described (P1), whilst in the[26]. five As output shown In the literature, it is possible to find several solutions for the design of a six-port junction. In our portsin Figure are connected 3, it has one respectively input port, tothe which reconfigurable is connected antenna to the (P2) power and amplifier to the power (P1), detectors whilst the (P3, five P4, case, considering the demand of a simple structure, we adopted the one described in [26]. As shown P5,output P6). ports are connected respectively to the reconfigurable antenna (P2) and to the power detectors in Figure 3, it has one input port, which is connected to the power amplifier (P1), whilst the five (P3, P4, P5, P6). output ports are connected respectively to the reconfigurable antenna (P2) and to the power detectors (P3, P4, P5, P6).

Electronics 2019, 8, 1191 4 of 16 ElectronicsElectronics 2019 2019, 8, ,8 x, xFOR FOR PEER PEER REVIEW REVIEW 4 4of of 16 16

FigureFigure 3. 3. Structure Structure of of the the six-port six-port junction. junction.

InIn Figure Figure3 3, ,3, thethe the six-portsix-port six-port junctionjunction junction isis is composed composed composed ofof of severalseveral several passivepassive passive elements:elements: elements: aa a 1212 12 dBdB dB directionaldirectional directional coupler,coupler, a a Wilkinson Wilkinson power power divider, divider, threethree three 9090° 90°◦ hybridhybr hybridid couplers, couplers, and and a a 3 3 dB dB attenuator. attenuator. The The choice choice of of thethe 12 12 dB dB coupling coupling factor factor (CF) (CF) for for the the directional directional coupler coupler was was made made as as a a tradeotradeoff tradeoffﬀ between between cost cost and and accuracyaccuracy of of the the power power detectable detectable at at the the four four output output ports. ports. Indeed, Indeed,Indeed, by by by increasing increasingincreasing the the the CF, CF, CF, the the calibration calibration algorithmalgorithm of of the the six-port six-port junction junction is is more more sensitive sensitive to to errors, errors, and and this this will will impact impact the the reading reading of of the the unknownunknown load load [[23,24]. 23[23,24].,24]. TheThe design designdesign was waswas considered considered over over a a awideband wideband wideband freque freque frequencyncyncy range range range starting starting starting from from from 5 5 to to 5 6 6 toGHz GHz 6 GHz [27] [27] [and 27and] andcarriedcarried carried outout outusingusing using aa commercial acommercial commercial CADCAD CAD environmentenvironment environment (i.e.,(i.e., (i.e., KeysightKeysight ADS ADSADS [28]); [28[28]);]); electromagneticelectromagneticelectromagnetic simulationssimulations (i.e., (i.e., Keysight Keysight Momentum, Momentum, embedded embedded in in Keysight Keysight ADS) ADS) were were extensively extensively adopted adopted to to increaseincrease the the accuracy accuracy ofof of the the design.design. design. TheThe boardboard waswas manufactured manufactured onon aa Rogers Rogers R04350BR04350 R04350BB low-costlow-cost PCBPCB substratesubstrate withwith aa dielectric dielectric constantconstant of of 3.48 3.48 and and a a thickness thickness of of 0.508 0.508 mm. mm. The The fabricated fabricated board board is is shown shown in in Figure Figure4 4.. 4.

FigureFigure 4. 4. Fabricated Fabricated six-port six-port junction. junction.

2.2.2.2. Detector Detector Design Design TheThe aim aim of of the thethe power powerpower detector detectordetector is isis to toto convert convert RF RF power power into into DC DC DC power power power [29]. [29]. [29 ].A A Apower power power detector detector detector is is isneededneeded needed at at the at the the six-port six-port six-port junction junction junction output output output ports ports ports dedicated dedicated dedicated to to sense sense to sense the the antenna theantenna antenna reflec reflec reflectiontiontion coefficient coefficient coefficient (i.e., (i.e., (i.e.,portsports ports P3, P3, P4, P4, P3, P5, P5, P4, and and P5, P6 andP6 in in P6Figure Figure in Figure 4). 4). The The4). classical classical The classical RF RF power power RF power detector detector detector is is composed composed is composed of of a a diode, diode, of a diode,a a low- low- R C apasspass low-pass filter filter (LPF) filter(LPF) (LPF)formed formed formed by by a a byload load a load resistor resistor resistor ( 𝑅(𝑅 ( Load) )and and) and a a capacitor acapacitor capacitor ( 𝐶( (𝐶Load),), ),and and an anan input inputinput matching matchingmatching networknetwork (IMN), (IMN), as as shown shown in in Figure Figure5 5.. 5.

ElectronicsElectronics 20192019,, 88,, 1191x FOR PEER REVIEW 55 of 1616 Electronics 2019, 8, x FOR PEER REVIEW 5 of 16

Figure 5. SimplifiedSimplified block scheme of the designed power detector. Figure 5. Simplified block scheme of the designed power detector. InIn thisthis work,work, wewe usedused aa SkyworksSkyworks SMS7630-079SMS7630-079 LFLF zero-biaszero-bias SchottkySchottky diodediode fromfrom (emission(emission coeffiIncient this n work,= 1, reverse we used saturation a Skyworks current SMS7630-079Is = 5 µA at LF 25 ).zero-bias Schottky diode from (emission coefficient n = 1, reverse saturation current Is = 5 μA at 25°).◦ coefficient n = 1, reverse saturation current Is = 5 μA at 25°). TheThe choice of R𝑅Load was was carried carried out out by by means means of of harmonic harmonicbalance balance simulationssimulations performedperformed withwith The choice of 𝑅 was carried out by means of harmonic balance simulations performed with KeysightKeysight ADS over thethe entireentire bandwidthbandwidth (from 55 toto 66 GHz).GHz). Since nono IMNIMN waswas designed,designed, thethe inputinput Keysight ADS over the entire bandwidth (from 5 to 6 GHz). Since no IMN was designed, the input powerpower hashas beenbeen sweptswept overover aa widewide range,range, i.e.,i.e., fromfrom −2020 toto ++2020 dBm, in order toto compensatecompensate forfor thethe power has been swept over a wide range, i.e., from −−20 to +20 dBm, in order to compensate for the inputinput mismatch.mismatch. input mismatch. TheThe resistanceresistanceR 𝑅Load has has been been selected selected to maximizeto maximize the the output output voltage voltage by sweeping by sweeping its value its value from The resistance 𝑅 has been selected to maximize the output voltage by sweeping its value 36fromΩ to36 36 Ωk toΩ .36 The kΩ upper. The valueupper was value chosen was chosen considering considering the quality the ofquality available of available off-the-shelf off-the-shelf resistors from 36 Ω to 36 kΩ. The upper value was chosen considering the quality of available off-the-shelf inresistors terms ofin parasiticterms of contributionsparasitic contributions in the entire in the frequency entire frequency band of interest. band of interest. resistors in terms of parasitic contributions in the entire frequency band of interest. AsAs shown in Figure6 6,, thethe outputoutput voltagevoltage isis directlydirectly proportionalproportional to the 𝑅RLoad value.value. FromFrom thisthis As shown in Figure 6, the output voltage is directly proportional to the 𝑅 value. From this plot,plot, we chose as best value R𝑅Load = =36 36k kΩΩsince since itit givesgives a higher value of the output-detected output-detected voltage. voltage. plot, we chose as best value 𝑅 = 36 kΩ since it gives a higher value of the output-detected voltage. ThisThis choicechoice willwill increaseincrease thethe accuracyaccuracy ofof thethe powerpower detection.detection. Using the results from Figure6 6,, we we can can This choice will increase the accuracy of the power detection. Using the results from Figure 6, we can alsoalso estimateestimate the saturation region of the detector,detector, whichwhich happenshappens approximatelyapproximately at 9 dBmdBm ofof inputinput also estimate the saturation region of the detector, which happens approximately at 9 dBm of input powerpower forfor thethe selectedselected impedance.impedance. This point, which will be affected affected by the input matching network, power for the selected impedance. This point, which will be affected by the input matching network, isis expectedexpected toto bebe beyondbeyond the the actual actual power power level level that that will will reach reach the the detector. detector. However, However, the the validation validation of is expected to be beyond the actual power level that will reach the detector. However, the validation theof the final final design design will will include include an an analysis analysis of of the the behavior behavior of of the the complete complete sensing sensing boardboard forfor realisticrealistic of the final design will include an analysis of the behavior of the complete sensing board for realistic powerpower levels,levels, toto verifyverify ifif saturationsaturation eeffectsffects arisearise fromfrom thethe detectordetector nonlinearity.nonlinearity. power levels, to verify if saturation effects arise from the detector nonlinearity.

FigureFigure 6.6. Output voltage versus input input power power with with frequency frequency sweep sweep from from 5 5 to to 6 6 GHz GHz when when 𝑅RLoad isis Figure 6. Output voltage versus input power with frequency sweep from 5 to 6 GHz when 𝑅 is 3636 kkΩΩ (green(green squaresquare line),line), 3.63.6 kkΩΩ (blue(blue triangletriangle line),line), andand 3636 ΩΩ (red(red solidsolid line).line). 36 kΩ (green square line), 3.6 kΩ (blue triangle line), and 36 Ω (red solid line). AfterAfter determiningdeterminingR𝑅 , the, the other other parameter parameter to be to addressed be addressed is C .is Its 𝐶 value. canIts bevalue determined can be After determining 𝑅Load , the other parameter to be addressedLoad is 𝐶 . Its value can be bydetermined looking at by the looking desired at bandwidth the desired of thebandwidth LPF. It is of defined the LPF. by the It is first-order defined LPFby the formed first-order by R LPFin determined by looking at the desired bandwidth of the LPF. It is defined by the first-orderLoad LPF parallelformed (consideringby 𝑅 in parallel the low-frequency (considering operation the low-frequency of the LPF) operation to the di ffoferential the LPF) diode to the resistance differentialR formed by 𝑅 in parallel (considering the low-frequency operation of the LPF) to the differentiald anddiode the resistance capacitor 𝑅 C [30 and,31 the]. capacitor C [30,31]. diode resistance 𝑅 and the capacitor C [30,31]. For this purpose, by looking at Equation (1), For this purpose, by looking at Equation (1), 1 𝑓 = 1 , 𝑅 =𝑅 ∥𝑅, (1) 𝑓 = 2𝜋𝑅𝐶, 𝑅 =𝑅 ∥𝑅, (1) 2𝜋𝑅𝐶

Electronics 2019, 8, 1191 6 of 16

For this purpose, by looking at Equation (1),

Electronics 2019, 8, x FOR PEER REVIEW 1 6 of 16 f3dB = , Rtot = Rd RLoad, (1) 2πRtotCLoad k it is possible to determine 𝐶 . Since the proposed reflectometer is designed to be cheap and itaccurate, is possible the idea to determine is to guaranteeCLoad. the Since possibility the proposed of analyzing reflectometer the antenna is designed impedance to bevariation cheap andin 1 accurate,GHz bandwidth the idea using is to narrowband guarantee the test possibility signals. For of example, analyzing the the entire antenna bandwidth impedance can be variation swept by in 1using GHz a bandwidth single-tone using signal narrowband over the grid test of signals. frequencies For example, for which the the entire calibration bandwidth has been can beperformed. swept by usingThis quickly a single-tone provides signal the overimpedance the grid of ofthe frequencies antenna, which for which can be the directly calibration used has to control been performed. the RMN Thisto optimize quickly the provides matching the impedancecondition. ofAs the a matter antenna, of whichfact, the can principle be directly is usedthe same to control of a spectrum the RMN toanalyzer optimize that the can matching explore very condition. large bandwidths, As a matter preserving of fact, thehigh principle accuracy is levels the sameand dynamic of a spectrum range, analyzerby exploiting that cannarrowband explore very IF filters. large bandwidths,To achieve this, preserving we design high the accuracy LPF to obtain levels anda bandwidth dynamic that range, is bysmall exploiting enough narrowbandto preserve the IF filters. accuracy To achieveof the power this, we detection, design thealthough LPF to not obtain too anarrow bandwidth to avoid that isexcessive small enough inertia toof preservethe filter, the which accuracy would of slow the powerthe frequency detection, sweep. although We chose not too 𝑓 narrow = 15 kHz to avoid as a excessivegood tradeoff, inertia for of which the filter, 𝑅 which = 36 k wouldΩ and slow 𝐶 the = 2.7 frequency nF are required. sweep. We chose f3dB = 15 kHz as a goodThe tradeo lastff step, for in which the detectorRLoad = 36 design kΩ and is relatedCLoad = to2.7 the nF IMN, are required. whose goal is to ensure the maximum powerThe transfer last step to inthe the diode. detector It has design to cover is related the entire to the frequency IMN, whose range goal (from is to 5 ensureto 6 GHz) the and maximum all the powerexpected transfer power to values the diode. (from It − has20 to to 0 coverdBm). the The entire power frequency range is rangenow reduced (from 5 tobecause 6 GHz) of and the allactual the expectedpower level power we expect values at (from the output20 to 0port dBm). of the The six-port power rangejunction is nowunder reduced actual becauseoperating of conditions the actual − power(i.e., input level power we expect of 10 atdBm). the output port of the six-port junction under actual operating conditions (i.e., inputTo our power best knowledge, of 10 dBm). it is the first time that the IMN of the detector is designed for 1 GHz of bandwidthTo our since, best knowledge, in typical applications it is the first [30,31], time that a narrowband the IMN of thedesign detector is adopted. is designed The IMN for 1 used GHz for of bandwidththis detector since, design in typicalis shown applications in Figure [730 and,31], makes a narrowband use of a design bridged is adopted.tee [32], an The RLC IMN matching used for thisnetwork, detector commonly design is shownused for in Figure broadband7 and makes cascadable use of again bridged stage tee design. [ 32], an RLCThis matchingkind of network,network commonlytransforms the used input for broadbandimpedance cascadableof the loaded gain detector stage design.diode toThis the required kind of network50 Ω system transforms impedance the inputby absorbing impedance its ofreactive the loaded part detectorinto an diodeall-pass to thefilter required over a 50 veryΩ system broad impedancefrequency byband absorbing [32]. The its reactiveinductors part (L1 intoand an L2 all-pass), the capacitor filter over (C) a and very the broad resistance frequency (R) bandvalues [32 of]. the The bridged inductors tee (L are1 and easily L2), thecalculated capacitor from (C) the and design the resistance equations (R) repo valuesrted of in the [32] bridged and their tee values are easily are: calculatedL1 = 0.36 nH, from L2 the= 0.44 design nH, equationsC = 0.080 pF, reported and R in= 50 [32 Ω] and. their values are: L1 = 0.36 nH, L2 = 0.44 nH, C = 0.080 pF, and R = 50 Ω.

Figure 7. Lumped-component bridged-tee network.

Due to the high-frequencyhigh-frequency application, the lumped components have been replaced with distributed ones designed in microstrip technology,technology, avoiding the unwanted parasitic effects effects of lumped SMD components. In In particular, particular, inductors are are realized with short transmission line length, and the series capacitor is implemented with an interdigitatedinterdigitated structure due to its small capacitance value. The 50 Ω resistor,resistor, instead, is a high frequency SMD flip-chipflip-chip thin-filmthin-film component.component. A broadband RF short circuit has been realized at the diode output by means of a radial stub to filterfilter out the residual fundamental frequencyfrequency andand thethe harmonicharmonic components.components. Each detector was designed considering the same Rogers substrate already used for the six-port junction (see(see Figure8 8).). TheThe detectordetector boardsboards havehave beenbeen fabricatedfabricated separatelyseparately fromfrom thethe six-portsix-port junctionjunction for testing purposes. Nevertheless, the use of the same substrate allows for easy integration of all the elements of the sensing board over thethe samesame PCB.PCB.

Electronics 2019, 8, 1191 7 of 16 Electronics 2019, 8, x FOR PEER REVIEW 7 of 16

Figure 8. Fabricated detector board. The The pad pad on the left has been included to ease the probing of the DC output.

2.3. Calibration Algorithm The most important issue in the design of thethe proposed sensing board is the calibration algorithm [26] [26] whose main aim is to derive the reflectionreflection coecoefficientfficient of the load cascaded to the sensing board (i.e., the antenna) from thethe voltagesvoltages providedprovided byby thethe detectors.detectors. In thethe literature,literature, it it is is possible possible to to find find plenty plenty of of algorithms algorithms dedicated dedicated to the to six-portthe six-port calibration calibration [26]. In[26]. this In work, this work, we adopted we adopted the five-standard the five-standard calibration calibration method, method, based based on a linear on a linear approach approach [33]. It [33]. can beIt can easily be easily implemented implemented in a common in a common programming programmi languageng language (e.g., C (e.g.,++, MATLAB),C++, MATLAB), and since and itsince is not it computationallyis not computationally expensive, expensive, it is suitable it is suitable for a simple for a simple IoT device. IoT device. The algorithmalgorithm consists consists of aof set a ofset mathematical of mathematical matrix matrix calculations. calculations. The main The step main is the step calculation is the ofcalculation the matrix ofC the, which matrix characterizes 𝑪, which characterizes the six-port the junction. six-port It junction. is a 4 4 It real is a matrix 4 × 4 real that matrix links thethat vector links × T T ofthe the vector four readof the powers fourP read= [p 3powers, p4, p5, p6𝑷] to = the𝑝,𝑝 unknown,𝑝,𝑝 load to vectorthe unknownΓ = [1, Γload , Re (vectorΓ), Im (𝜞Γ)] = according 1, |𝛤|,𝑅𝑒 to𝛤 the,𝐼𝑚 following𝛤 according relationship to the [ 26following,33]: relationship [26,33]: 𝑷=𝜌∙𝑪∙𝚪, (2) P = ρ C Γ, (2) · · where 𝜌=|𝑏| is the incident wave power at the antenna input port (i.e., antenna in Figure 2). where ρ = b 2 is the incident wave power at the antenna input port (i.e., antenna in Figure2). The calibration| 2| of the six-port junction is based on the identification of 𝑪 by means of the measurementThe calibration of the vector of the 𝑷 six-port for five junction known loads is based used on as the a re identificationference. Then, of theC genericby means unknown of the measurementload can be derived of the by vector usingP thefor matrix five known 𝑿=𝑪 loads𝟏 according used as ato reference. the following Then, equations the generic [26]: unknown 1 load can be derived by using the matrix X = C− according to the following equations [26]: ∑ ∑ 𝑅𝑒 Γ = 𝐼𝑚 Γ = , (3) P∑4 P∑4 j=1 x3jPj+2 j=1 x4jPj+2 Re(Γ) = Im(Γ) = , (3) where 𝑃 is the power value at portP j 4when measuring the unknownP4 load, and 𝑥 are the elements j= x1jPj+2 j= x1jPj+2 of 𝑿. 1 1 whereThePj fiveis the calibration power value loads at portshouldj when be chosen measuring in order the unknown to maximize load, the and accuracyxij are the and elements the stability of X. of theThe algorithm. five calibration In our practical loads should case, bewe chosen chose an in orderexperimental to maximize approach the accuracy for the selection and the stabilityof these ofloads. the algorithm.Taking into In account our practical the application case, we chose for th ane sensing experimental board, approach we considered for the selectionsome impedance of these loads.measurements Taking intoof a accountdedicated the 5.15–5.85 application GHz for antenna the sensing array for board, high we data considered rate ultra-short-range some impedance 3 × 3 measurementsmultiple input multiple of a dedicated output 5.15–5.85 (MIMO) GHz wireless antenna communications array for high [34]. data The rate antenna ultra-short-range impedance 3 was3 × multiplemeasured input in three multiple different output configurations, (MIMO) wireless i.e., as: communications a standalone antenna, [34]. The integrated antenna impedance in a worktop, was measuredand mounted in three underneath different the configurations, worktop. The i.e., meas as:urement a standalone results antenna, are shown integrated in Figure in a 9. worktop, and mounted underneath the worktop. The measurement results are shown in Figure9.

Electronics 2019, 8, 1191 8 of 16 Electronics 2019, 8, x FOR PEER REVIEW 8 of 16

Figure 9. Antenna impedance measurements for three diﬀfferenterent scenarios:scenarios: standalone antenna (red solid line) integrated in a worktop (green reverse triangle line) and mounted underneath the worktop (blue thickthick line).line). Frequency goes from 5 to 66 GHz accordingaccording toto thethe directiondirection ofof thethe arrow.arrow.

Starting from these three possible conditions, we chose fivefive loads to optimize the accuracy of the sensing-board algorithm in the area defined defined by the impedances reported in Figure 99 overover thethe entireentire bandwidth. In addition,addition, wewe tooktook into into account account the the condition condition number number of of the the matrix matrixC [35𝑪 ] [35] to reduce to reduce the sensitivitythe sensitivity of the of calibration the calibration algorithm algorithm to measurement to measurement uncertainty. uncertainty. The selected The loads selected are reported loads are in Tablereported1, for in which Table the1, for condition which the number condition results number su fficiently results smallsufficiently (~5). small (~5).

Table 1.1. Calibration Standards (ΩΩ).).

11 22 33 44 5 5 43.5743.57+ +j j 11.45 11.45 0.821 0.821+ +j j 98.67898.678 65 65 0.030.03 – jj 25.7 25.7 21.8 21.8 – j 39.68j 39.68 − − Matrix 𝑪 (and, therefore, matrix 𝑿) must be calculated for each frequency of interest. This is a Matrix C (and, therefore, matrix X) must be calculated for each frequency of interest. This is time-consuming operation, especially if a small frequency resolution is required over the entire a time-consuming operation, especially if a small frequency resolution is required over the entire design bandwidth. Moreover, it requires the availability of the five “standards” designed over a wide design bandwidth. Moreover, it requires the availability of the five “standards” designed over a wide frequency range. frequency range. We propose a different approach which is sufficiently accurate and less time consuming, since We propose a different approach which is sufficiently accurate and less time consuming, since it is implemented via simulation, e.g., by using Keysight ADS. We reproduce the response of the it is implemented via simulation, e.g., by using Keysight ADS. We reproduce the response of the sensing board modeled by the measured S-parameters for the six-port junction and the sensing board modeled by the measured S-parameters for the six-port junction and the electromagnetic electromagnetic simulation together with the detector diode nonlinear model. We connect at port 2 simulation together with the detector diode nonlinear model. We connect at port 2 (P2, Figure4) each (P2, Figure 4) each calibration loads for the all frequencies of interest (e.g., from 5 to 6 GHz with a calibration loads for the all frequencies of interest (e.g., from 5 to 6 GHz with a step of 100 MHz) to step of 100 MHz) to obtain the four output voltages. Then, matrix 𝑪 and the calibration algorithm obtain the four output voltages. Then, matrix C and the calibration algorithm are implemented via are implemented via MATLAB starting from the simulation results. Apart from the time reduction MATLABcompared startingwith the from long the measurement simulation results.process, Apart one ad fromditional the timeadvantage reduction is the compared possibility with to perform the long measurementthis procedure process, without one the additional need for real advantage dedicated is the calibration possibility loads to perform or equivalent this procedure elements without (e.g., a thepassive need tuner) for real for dedicatedthe synthesis calibration of the required loads orimpedances, equivalent which elements reduces (e.g., the a passivecost of the tuner) procedure. for the synthesisUsing of such the required an implementation, impedances, whichthe calibration reduces theprocedure cost of the that procedure. leads to matrix 𝑪 should be C performedUsing only such once, an implementation, and then the result the can calibration be stored procedure as matrix that𝑿=𝑪 leads𝟏 to to directly matrix useshould Equation be X = C 1 performed(3) for the onlyimpedance once, and detection. then the Since result this can matrix be stored is composed as matrix of 16 −realto directlyfloating-point use Equation numbers, (3) forassuming the impedance a 32-bit representation, detection. Since only this matrix64 bytes is of composed memory ofare 16 required real floating-point for each calibrated numbers, frequency. assuming aFrom 32-bit a computational representation, point only of 64 view, bytes Equation of memory (3) areonly required requires for 16 eachmultiplications, calibrated frequency. 12 additions, From and a computational2 divisions for pointeach frequency of view, Equation we want (3) to only detect requires the reflection 16 multiplications, coefficient. 12 Such additions, a limited and number 2 divisions of foroperations each frequency and memory we want torequirements detect the reflection are today coeffi cient.largely Such satisfied a limited by number common of operations low-cost microcontrollers, which can be suitable for the applications we are addressing.

Electronics 2019, 8, 1191 9 of 16 and memory requirements are today largely satisﬁed by common low-cost microcontrollers, which can beElectronics suitable 2019 for, the8, x FOR applications PEER REVIEW we are addressing. 9 of 16 Considering Equation (3), the power needs to be calculated from the detector output voltages. Considering Equation (3), the power needs to be calculated from the detector output voltages. To this purpose, an analysis was done to evaluate the nonlinearity of the detector (see Section 3.3.1). To this purpose, an analysis was done to evaluate the nonlinearity of the detector (see Section 3.3.1). 3. Measurement Section 3. Measurement Section In this section, the measurements for the test and validation of the sensing board are reported. In this section, the measurements for the test and validation of the sensing board are reported. 3.1. Validation of the Fabricated Boards 3.1. Validation of the Fabricated Boards S-parameter measurements have been carried out to validate the accuracy of the six-port S-parameter measurements have been carried out to validate the accuracy of the six-port junction [27]. We performed them with a Keysight PNA-X in the bandwidth from 4 to 7 GHz, junction [27]. We performed them with a Keysight PNA-X in the bandwidth from 4 to 7 GHz, and and reported the results in Figure 10. From [23,24], the paths between P3–P1 and P6–P1 and between reported the results in Figure 10. From [23,24], the paths between P3–P1 and P6–P1 and between P5– P5–P1P1 and and P4–P1 P4–P1 have have to to be be equal, equal, and and these these requir requirementsements are are approximately approximately satisfied satisﬁed over over the the whole whole bandwidth,bandwidth, notably notably in in the the targeted targeted frequencyfrequency rangerange (5–6 GHz). In In Figure Figure 10, 10 we, we also also report report the the insertioninsertion loss loss of theof the six-port six-port junction, junction, which which is approximatelyis approximately constant constant and and equal equal to 1to dB 1 dB over over the the band of interest.band of interest.

FigureFigure 10. 10.S-parameter S-parameter measurements measurements ofof thethe six-portsix-port junction.junction. Upper Upper row: row: the the magnitude magnitude of ofS31 S31 and and S41S41 is reportedis reported with with red red solid solid line line and andthe the magnitudemagnitude of S61 and and S51 S51 is is reported reported with with thick thick blue blue solid solid line.line. The The phase phase of S31of S31 and and S41 S41 is reported is reported with with green green dotted dotted line line and and the phasethe phase of S61 of andS61 S51and isS51 shown is withshown black with square black line. square Lower line. row: Lower insertion row: insertion loss. loss.

TheThe return return loss loss of of the the fabricated fabricated detectordetector prototypesprototypes was was measured measured at at different diﬀerent power power levels levels to to verifyverify the the eﬀ effectivenessectiveness of of the the matching matching networknetwork as a function of of power power over over the the entire entire bandwidth. bandwidth. TheThe results results reported reported in in Figure Figure 11 11 show show aa returnreturn lossloss of −1515 dB dB or or less, less, close close to to the the central central frequency frequency − (5.5(5.5 GHz). GHz). It It is is noteworthy noteworthy how how the the designeddesigned matchingmatching network network strongly strongly redu reducesces the the variation variation of ofthe the return loss over a considerable range of input power. return loss over a considerable range of input power.

Electronics 2019,, 88,, 1191x FOR PEER REVIEW 1010 of of 16 Electronics 2019, 8, x FOR PEER REVIEW 10 of 16

Figure 11. Measurement results of the return loss of the detector for different input power levels. Red Figuresolid line 11. is Measurement 0 dBm, blue thick results line of is the −5 returndBm, green loss of reverse the detector triangle for line didifferentff erentis −10 inputinput dBm, power powerviolet levels.squarelevels. Redline solidis −20 lineline dBm. isis 00 dBm,dBm, blueblue thickthick lineline isis −5 dBm, green reverse triangle lineline isis −10 dBm, violet square line − − isis −20 dBm. − 3.2. Validation of the Impedance Detection 3.2. Validation of the Impedance Detection In this section, a validation of the designed sensing board as a reflectometer is performed with In this section, a validation of the designed sensing board as a reflectometer is performed with the setupIn this shown section, in Figurea validation 12, which of the includes designed an seRFnsing microwave board as source a reflectometer (Keysight N5182B),is performed a Maury with the setup shown in Figure 12, which includes an RF microwave source (Keysight N5182B), a Maury microwavethe setup shown mechanical in Figure tuner, 12, andwhich a digital includes multimet an RF microwaveer (Rigol DM3058). source (Keysight By means N5182B), of a set of a single-Maury microwave mechanical tuner, and a digital multimeter (Rigol DM3058). By means of a set of single-tone tonemicrowave measurements, mechanical from tuner, 5 to and6 GHz a digital with 100 multimet MHz stereps, (Rigol we DM3058).are able to By test means the sensing of a set board of single- (six- measurements, from 5 to 6 GHz with 100 MHz steps, we are able to test the sensing board (six-port porttone junctionmeasurements, + detector) from over 5 to the 6 GHz entire with bandwidt 100 MHzh of st interest.eps, we Theseare able measurements to test the sensing were carriedboard (six- out junctionbyport applying junction+ detector) one+ detector) tone over for over theeach entirethe frequency entire bandwidth bandwidt with ofa hconstant interest. of interest. input These These power measurements measurements of 10 dBm were (8.27were carried dBmcarried outat outthe by applyingsensingby applying board one one tone input tone for port) eachfor andeach frequency settling frequency with the impedance awith constant a constant input values powerinput by a powermechanical of 10 dBm of 10 (8.27 tunerdBm dBm in(8.27 order at dBm the to sensing avoidat the boardanysensing change input board due port) input to and theport) settling interactions and settling the impedance with the impedancethe en valuesvironment values by a mechanical bythat a mechanicalwould tuner be produced tuner in order in order toby avoidan to actualavoid any changeantenna.any change due These todue the impedances to interactions the interactions were with measured thewith environment the in en advancevironment that by would meansthat bewould of produced a vectorbe produced bynetwork an actual by analyzer an antenna. actual to Thesegatherantenna. impedancestheir These values impedances werewith measureda high were level measured in of advance accuracy, in by advance so means that theyofby a means vectorcan be network ofconsidered a vector analyzer networka good to reference gatheranalyzer their for to valuesthegather validation withtheir avalues high of the level with sensing of a accuracy,high board. level so of thatThe accuracy, theyread can vo soltages be that considered theyfrom can the a be goodmultimeter considered reference were a for good theprocessed reference validation and, for of thethrough sensingvalidation the board. calibration of the The sensing read algorithm voltages board. explained from The theread multimeter in vo Sectltagesion werefrom 2.3, processedthethe loadmultimeter reflection and, through were coefficients processed the calibration wereand, algorithmderived.through Thesethe explained calibration results in are Sectionalgorithm reported 2.3, explainedin the Table load 2 reflection inand Sect in ionFigure coe 2.3,ffi cients13. the We load were can reflection derived.conclude Thesecoefficientsthat the results detected were are reportedimpedancesderived. These in Tableare results in2 goodand are inagreement Figurereported 13 within. WeTable the can ones2 concludeand se int byFigure thatthe mechanical the13. We detected can tunerconclude impedances over that the areinvestigatedthe indetected good agreementbandwidth.impedances with are thein good ones agreement set by the mechanical with the ones tuner set over by the the mechanical investigated tuner bandwidth. over the investigated bandwidth.

Figure 12. MeasurementMeasurement setup for the validation of the the se sensingnsing board board with with single-tone single-tone measurements. Figure 12. Measurement setup for the validation of the sensing board with single-tone measurements.

Electronics 2019, 8, x FOR PEER REVIEW 11 of 16 Electronics 2019, 8, 1191 11 of 16

Table 2. Impedance Measurement Results.

Table 2. Impedance Measurement Results.Retrieved Values Frequency (GHz) Tuner Impedances (Ω) from the sensing board (Ω) Frequency (GHz) Tuner Impedances (Ω) Retrieved Values from the Sensing Board (Ω) 5 43.6 + j 12.04 40.3 + j 12.0 55.1 43.6 + j 12.04 30.5 + j 1.32 40.3 34.9+ j 12.0+ j 7.5 5.1 30.5 + j 1.32 34.9 + j 7.5 5.2 48.1 – j 12.1 44.4 – j 6.0 5.2 48.1 j 12.1 44.4 j 6.0 − − 5.35.3 48.6 + j 48.6 4.2 + j 4.2 45.7 45.7+ j + 2.5 j 2.5 5.45.4 30.9 + j 30.9 3.9 + j 3.9 36.6 36.6+ j + 6.6 j 6.6 5.55.5 45.9 j45.9 8.8 – j 8.8 44.7 44.7j – 10.9 j 10.9 − − 5.65.6 38.4 + j 38.4 16.9 + j 16.9 40.7 40.7+ j + 17.3 j 17.3 5.7 32.5 j 1.9 33.2 j 2.4 5.7 − 32.5 – j 1.9 33.2− – j 2.4 5.8 54.4 j 7.4 52.5 j 5.5 5.8 + 54.4 + j 7.4 52.5+ + j 5.5 5.9 37.2 + j 10.3 38.2 + j 12.4 5.9 37.2 + j 10.3 38.2 + j 12.4 6 49.4 + j 4.8 49.9 + j 1.0 6 49.4 + j 4.8 49.9 + j 1.0

FigureFigure 13. 13.Measurement Measurement results results are are reported. reported. Red Red circle circle line line shows showsthe the reflectionreflection coecoefficientfficient inin termsterms of realof and real imaginaryand imaginary part part set at set the at tuner.the tuner. Blue Blue solid solid line line shows shows the the reflection reflection coe coefficientfficient in in terms terms of of real andreal imaginary and imaginary part retrieved part retrie fromved thefrom sensing the sensing board. board.

3.3.3.3. Sensing Sensing Board Board Validation Validation with with Large-Signal Large-Signal MeasurementsMeasurements InIn this this section, section, a a nonlinearity nonlinearity analysis analysis ofof thethe sensingsensing board and and a a bandwidth bandwidth estimation estimation of ofthe the detectorsdetectors are are carried carried out. out. To To perform perform both both analyses,analyses, twotwo didifferentﬀerent measurement setups setups were were used. used.

3.3.1.3.3.1. Nonlinearity Nonlinearity of of Detectors Detectors and and Sensing Sensing BoardBoard LookingLooking at at Equation Equation (2), (2), we we needneed toto knowknow the power at at the the four four output output ports ports of of the the six-port six-port junctionjunction in orderin order to retrieve to retrieve the information the information about about the load the reflection load reflection coefficient. coefficient. They can They be measuredcan be throughmeasured a power through meter a power as in [ 27meter] or withas in the[27] setupor with implemented the setup implemented in [36], which in [36], basically which works basically as the powerworks detectors. as the power detectors. Considering the designed sensing board, we should retrieve the power values from the output Considering the designed sensing board, we should retrieve the power values from the output voltages of the detectors. The direct use of the well-known proportionality between power and voltages of the detectors. The direct use of the well-known proportionality between power and squared squared voltage (𝑃∝𝑉) is correct only from a theoretical perspective. In practice, the nonlinearity voltage (P V2) is correct only from a theoretical perspective. In practice, the nonlinearity of the of the detectors∝ can alter this relationship, causing a poor level of accuracy in the determination of detectors can alter this relationship, causing a poor level of accuracy in the determination of the the load impedance. As a consequence, the 𝑃 −𝑉 characteristic of the detectors has to be load impedance. As a consequence, the P Vout characteristic of the detectors has to be properly properly characterized in order to obtain accurain − te results over a wide range of power values. characterizedIt should in be order pointed to obtain out that accurate a nonlinear results verification over a wide of rangethe detectors of power is required values. even if the RF front-endIt should is beintended pointed to out operate that a at nonlinear constant verification power. Indeed, of the the detectors power islevel required reaching even the if thefour RF front-enddetectors is intendedof the sensing to operate board atis constantnot the same power. for every Indeed, port the of power the six-port level reachingjunction theand, four in general, detectors of thedepends sensing on boardthe actual is not front-end the same load. for every port of the six-port junction and, in general, depends on the actualBy front-endcharacterizing load. the detectors 𝑃 −𝑉 relationship, we can extend the sensing board operationBy characterizing over a considerably the detectors wideP rangeVout of relationship,power. To this we end, can we extend used the the sensing setup in board Figure operation 14. A in − overKeysight a considerably PNA-X, in wide nonlinear range vector of power. network To this analyzer end, we (NVNA) used the mode, setup has in been Figure used 14 .to A perform Keysight PNA-X, in nonlinear vector network analyzer (NVNA) mode, has been used to perform nonlinear measurements on the sensing board with an input power sweep from 10 to 25 dBm. To obtain the − Electronics 2019, 8, x FOR PEER REVIEW 12 of 16 Electronics 2019, 8, 1191 12 of 16 nonlinearElectronics 2019measurements, 8, x FOR PEER on REVIEW the sensing board with an input power sweep from −10 to 2512 dBm. of 16 To obtain the highest levels of power, it was necessary to introduce a power amplifier (frequency range highest2 tononlinear 8 GHz, levels measurementsgain of power, 35 dBm, it was 𝑃on the necessary sensing = 34 dBm). toboard introduce Through with an a two powerinput directional power amplifier sweep couplers, (frequency from one−10 range toat 25the 2dBm. input to 8 GHz,To and gainoneobtain at 35 the dBm, the output, highestP1dBm it levels =is 34possible dBm).of power, to Through retrieve it was necessary twothe incident directional to introduce and couplers, reflected a power onewaveforms amplifier at the input at (frequency the and input one range and at thethe output,output2 to 8 itportsGHz, is possible gainof the 35 to dBm,sensing retrieve 𝑃 board the = incident 34 from dBm). andwhich Through reflected the two actual waveforms directional input at powercouplers, the input and one and the at the theload outputinput reflection and ports ofcoefficientone the sensingat the can output, boardbe calculated. it fromis possible which The to voltages theretrieve actual theat inputthe incident fo powerur outputand andreflected ports the loadofwaveforms the reflection sensing at the board coe inputffi cientare and detected can the be calculated.byoutput means ports of The two voltagesof multimetersthe sensing at the fourboard(Rigol output from DM3058 portswhich and ofthe theHewlett actual sensing inputPackard board power are34401A). detectedand the As byloadload, means reflection we ofused two a multimetersmechanicalcoefficient tuner. (Rigolcan be DM3058calculated. and The Hewlett voltages Packard at the fo 34401A).ur output As ports load, of wethe usedsensing a mechanical board are detected tuner. by means of two multimeters (Rigol DM3058 and Hewlett Packard 34401A). As load, we used a mechanical tuner.

FigureFigure 14.14. MeasurementMeasurement setupsetup forfor detectordetector nonlinearitynonlinearity investigation.investigation. ItIt includesincludes thethe sensingsensing boardboard Figure 14. Measurement setup for detector nonlinearity investigation. It includes the sensing board (six-port(six-port junction junction and and detectors) detectors) and and the variablethe variable load implementedload implemente withd a mechanicalwith a mechanical tuner; the tuner; voltages the (six-port junction and detectors) and the variable load implemented with a mechanical tuner; the arevoltages detected are bydetected two multimeters by two multimeters and the signal and the is applied signal is by applied the PNA-X. by the PNA-X. voltages are detected by two multimeters and the signal is applied by the PNA-X. TheThe resultsresults fromfrom thesethese measurementsmeasurements areare reportedreported inin FigureFigure 15 15.. Here,Here, thethe frequencyfrequency isis constantconstant The results from these measurements are reported in Figure 15. Here, the frequency is constant atat 5.55.5 GHz,GHz, thethe inputinput powerpower sweepssweeps fromfrom −1010 toto 2525 dBmdBm andand thethe impedanceimpedance setset byby thethe tunertuner isis at 5.5 GHz, the input power sweeps from −−10 to 25 dBm and the impedance set by the tuner is Z𝑍target==63 63 −j 19 𝑗 Ω19 . We𝛺. We used used the thePin 𝑃V−𝑉out characteristic characteristic of the of detectors the detectors to estimate to estimate the actual the poweractual 𝑍 =− 63 − 𝑗 19 𝛺. We used the −𝑃 −𝑉 characteristic of the detectors to estimate the actual power delivered by the four output ports of the six-port junction. deliveredpower delivered by the four by outputthe four ports output of ports the six-port of the six-port junction. junction.

FigureFigureFigure 15.15. 15.Real Real Real and and and imaginary imaginary imaginary part partpart of of a detected a detected impedance impedaimpedancence as aas as function a afunction function of theof of the input the input input power power power at 5.5at GHz.5.5at 5.5 GHz.GHz. As expected, the detected impedance is approximately constant over the whole range of input As expected, the detected impedance is approximately constant over the whole range of input power,As and expected, data results the detected agree withimpedance the value is approxim set by theately tuner. constant This conﬁrms over the thewhole usefulness range of of input the power, and data results agree with the value set by the tuner. This confirms the usefulness of the introductionpower, and data of the results detector agreePin withVout theactual value characteristic set by the tuner. in the calibrationThis confirms algorithm the usefulness of the sensing of the introduction of the detector 𝑃 −−𝑉 actual characteristic in the calibration algorithm of the sensing boardintroduction to extend of the its functionalitydetector 𝑃 −𝑉 to a wide actual range characteristic of power. in the calibration algorithm of the sensing board to extend its functionality to a wide range of power. board to extend its functionality to a wide range of power.

Electronics 2019, 8, 1191 13 of 16

Electronics 2019, 8, x FOR PEER REVIEW 13 of 16 Electronics 2019, 8, x FOR PEER REVIEW 13 of 16 3.3.2.3.3.2. Bandwidth Bandwidth of of the the Detector Detector 3.3.2. Bandwidth of the Detector InIn this this section, section, we we verify verify the the bandwidth bandwidth ofof thethe detector.detector. For For this this purpose, purpose, the the measurement measurement conditionconditionIn this must must section, be be properly properly we verify defined.defined. the bandwidth The The low-pass low-pass of thebehavior behavior detector. of ofthe For the detector this detector purpose, cannot cannot bethe investigated bemeasurement investigated by byconditionusing using single-tone single-tone must be measurements. properly measurements. defined. In fact, InThe fact, thelow-pass the5 to 56 toGHzbehavior 6 GHz RF IMN of RF the IMNprevents detector prevents low-frequency cannot low-frequency be investigated signals signalsfrom by fromusingflowing flowing single-tone through through themeasurements. thecircuit, circuit, whereas whereas In fact, an RFthe an sinusoid5 RF to sinusoid6 GHz will RF willprovide IMN provide prevents only onlya constantlow-frequency a constant output output signals voltage. voltage.from To Toflowingsolve this this through problem, problem, the the circuit, the idea idea was whereas was to use to an the use RF detector thesinusoid detector to willreveal to provide revealthe envelope only the envelopea ofconstant a 2-toneof output signal a 2-tone voltage. by varying signal Toby varyingsolvethe frequency this the problem, frequency of the the envelope of idea the envelopewas itself to use toitself thecharacteri detector to characterizeze to the reveal frequency the envelope frequency response of responseaof 2-tone the output signal of the LPF. by output varying LPF. theWe frequencyWe used used the ofthe setup the setup envelope shown shown itself in in Figure Figure to characteri 16 16,, wherewhereze the anan frequency RF source response(Keysight (Keysight of N5182B) N5182B)the output is isused LPF. used to toset set a a two-tonetwo-toneWe signal used signal atthe at the setup the input input shown of of the the in detector Figuredetector 16, board,board, where and an an RF oscillosc oscilloscopesource (Keysightope (Tektronix (Tektronix N5182B) DPO4104) DPO4104) is used acquires to acquires set a two-tone signal at the input of the detector board, and an oscilloscope (Tektronix DPO4104) acquires thethe output output waveform. waveform. For For the the 2-tone 2-tone signal signal generation, generation, we we consider consider the the central central frequency frequencyf0 of𝑓 5.5of 5.5 GHz the output waveform. For the 2-tone signal generation, we consider the central frequency 𝑓 of 5.5 andGHz we and swept we the swept distance the distance∆ f between Δ𝑓 between the two the tones two fromtones 200from Hz 200 to Hz 40 kHz.to 40 kHz. GHz and we swept the distance Δ𝑓 between the two tones from 200 Hz to 40 kHz. ToTo evaluate evaluate the the real real bandwidth bandwidth of of thethe detector,detector, thethe output waveform waveform is is processed processed through through a a To evaluate the real bandwidth of the detector, the output waveform is processed through a mathematicalmathematical procedure procedure to to calculate calculate thethe peak-to-peakpeak-to-peak voltage for for each each value value of of Δ𝑓∆.f The. The results results are are mathematical procedure to calculate the peak-to-peak voltage for each value of Δ𝑓. The results are reportedreported in in Figure Figure 17 17.. The The cuto cutoffff frequency frequency ( (−33 dBdB point) is approximately approximately 13.4 13.4 kHz, kHz, which which is isin ingood good reportedagreement in withFigure our 17. design The cutoff specification frequency (i.e., (−−3 15 dB kHz). point) is approximately 13.4 kHz, which is in good agreementagreement with with our our design design specification specification (i.e., (i.e., 1515 kHz).kHz).

Figure 16. Measurement setup for the investigation of the bandwidth of the detectors. An RF source FigureFigureis used 16. 16. Measurementto Measurementset the two-tone setup setup forsignal for the the and investigation investigation the oscilloscope of of the the bandwidth bandwidthis needed oftoof thetheacquire detectors.detectors. the output An RF voltagesource is usediswaveform. used to set to the set two-tone the two-tone signal andsignal the and oscilloscope the oscilloscope is needed is toneeded acquire to theacquire output the voltage output waveform. voltage waveform.

FigureFigure 17. 17.Peak-to-peak Peak-to-peak voltage voltage of of the the outputoutput waveformwaveform pr providedovided by by the the detector detector as as a afunction function of ofthe the Figure 17. Peak-to-peak voltage of the output waveform provided by the detector as a function of the frequencyfrequency spacing spacing∆ fΔ𝑓. . frequency spacing Δ𝑓.

Electronics 2019, 8, 1191 14 of 16

4. Conclusions In this work, we have described an ultra-wideband sensing board for IoT applications. We have extensively characterized the sensing board by means of small- and large-signal measurements in the frequency range of interest (i.e., 5 to 6 GHz) to deeply verify the goodness of the proposed design approach. In particular, we have also investigated the sensing board nonlinear behavior and veriﬁed the possibility of extending its application to a wide range of power levels.

Author Contributions: Investigation, A.P. and G.B.; Methodology, A.R., G.A. and D.R.; Supervision, A.R., G.V. and D.S.; Writing—original draft, A.P. and G.B.; Writing—review & editing, A.R., D.R., G.A., G.V. and D.S. Funding: This research was funded in part by the Eurostars project E!10149 MicromodGaN. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Bandyopadhyay, D.; Sen, J. Internet of Things: Applications and Challenges in Technology and Standardization. Wirel. Pers. Commun. 2011, 58, 49–69. [CrossRef] 2. Jo, M.; Maksymyuk, T.; Batista, R.L.; Maciel, T.F.; de Almeida, A.L.F.; Klymash, M. A survey of converging solutions for heterogeneous mobile networks. IEEE Wirel. Commun. 2014, 21, 54–62. 3. Minoli, D.; Occhiogrosso, B. Ultrawideband (UWB) Technology for Smart Cities IoT Applications. In Proceedings of the IEEE International Smart Cities Conference, Kansas City, MO, USA, 16–19 September 2018; pp. 1–8. 4. ETSI EN 301 893. 5 GHz RLAN; Harmonised Standard Covering the Essential Requirements of Article 3.2 of Directive 2014/53/EU; ETSI EN: Sophia Antipolis, France, 2014. 5. Huang, C.P.; Christainsen, K.; Nabokin, S.; Mirzayantz, R.; Allum, J.; Chen, A.; Lam, L.; McPartlin, M.; Doherty, M.; Vaillancourt, B. A highly integrated single chip 5–6 GHz front-end IC based on SiGe BiCMOS that enhances 802.11ac WLAN radio front-end designs. In Proceedings of the IEEE Radio Frequency Integrated Circuits Symposium (RFIC), Phoenix, AZ, USA, 17–19 May 2015; pp. 227–230. 6. Salpietro, R.; Bedogni, L.; di Felice, M.; Bononi, L. Park Here! a smart parking system based on smartphones’ embedded sensors and short range Communication Technologies. In Proceedings of the IEEE 2nd World Forum on Internet of Things (WF-IoT), Milan, Italy, 14–16 December 2015; pp. 18–23. 7. Lukić,M.; Mihajlović,Ž.; Mezei, I. Data Flow in Low-Power Wide-Area IoT Applications. In Proceedings of the 26th Telecommunications Forum (TELFOR), Belgrade, Serbia, 21–22 November 2018; pp. 1–4. 8. Pal, A.; Kant, K. NFMI: Connectivity for Short-Range IoT Applications. Computer 2019, 52, 63–67. [CrossRef] 9. ETSI EN 302 065-5. Short Range Devices (SRD) Using Ultra Wide Band Technology (UWB); Harmonised Standard Covering the Essential Requirements of Article 3.2 of Directive 2014/53/EU; Part 5: Devices using UWB technology onboard aircraft; ETSI EN: Sophia Antipolis, France, 2014. 10. Ramos, A.; Lazaro, A.; Girbau, D.; Villarino, R. RFID and Wireless Sensors Using Ultra-Wideband Technology; ISTE Press—Elsevier: London, UK, 2016; pp. 14–16. 11. Raffo, A.; di Falco, S.; Sozzi, G.; Menozzi, R.; Schreurs, D.M.M.-P.; Vannini, G. Analysis of the gate current as a suitable indicator for FET degradation under nonlinear dynamic regime. Microelectron. Reliab. 2011, 51, 235–239. [CrossRef] 12. Vadalà, V.; Bosi, G.; Raffo, A.; Vannini, G.; Avolio, G.; Schreurs, D.M.P. Influence of the gate current dynamic behaviour on GaAs HEMT reliability issues. In Proceedings of the 7th European Microwave Integrated Circuit Conference, Amsterdam, The Netherlands, 29–30 October 2012; pp. 258–261. 13. Pozar, D.M. Microwave Engineering, Reading; Addison-Wesley: Boston, MA, USA, 1990; pp. 529–565. 14. Mohamed, A.M.M.; Boumaiza, S.; Mansour, R.R. Novel Reconfigurable Fundamental/Harmonic Matching Network for Enhancing the Efficiency of Power Amplifier. In Proceedings of the 40th European Microwave Conference, Paris, France, 28–30 September 2010; pp. 1122–1125. 15. Nawaz, A.A.; Albrecht, J.D.; Ulusoy, A.C. A Ka-Band Power Amplifier with Reconfigurable Impedance Matching Network. In Proceedings of the IEEE 19th Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems (SiRF), Orlando, FL, USA, 20–23 January 2019; pp. 1–3. Electronics 2019, 8, 1191 15 of 16

16. Chen, H.; Hsieh, Y.; Lu, L. A 5.5-GHz multi-mode power amplifier with reconfigurable output matching network. In Proceedings of the IEEE Radio Frequency Integrated Circuits Symposium (RFIC), Phoenix, AZ, USA, 17–19 May 2015; pp. 203–206. 17. Lu, Y.; Peroulis, D.; Mohammadi, S.; Katehi, L.P.B. A MEMS reconfigurable matching network for a class AB amplifier. IEEE Microw. Wirel. Compon. Lett. 2003, 13, 437–439. 18. Baroni, A.; Nepa, P.; Rogier, H. Wearable self-tuning antenna for emergency rescue operations. IET Microw. Antennas Propag. 2016, 10, 173–183. [CrossRef] 19. Hur, B.; Eisenstadt, W.R.; Melde, K.L. Testing and Validation of Adaptive Impedance Matching System for Broadband Antenna. Electronics 2019, 8, 1055. [CrossRef] 20. Ji, D.; Jeon, J.; Kim, J. A Novel Load Insensitive RF Power Amplifier Using a Load Mismatch Detection and Curing Technique. In Proceedings of the 2013 IEEE Radio Frequency Integrated Circuits Symposium (RFIC), Seattle, WA, USA, 2–4 June 2013; pp. 341–344. 21. Curticipean, F.; Niittylahti, J. An Improved Digital Quadrature Frequency Down-Converter Architecture. In Proceedings of the 1977 11th Asilomar Conference on Circuits, Systems and Computers, Pacific Grove, CA, USA, 7–9 November 1977; Conference Record; IEEE: Piscataway, NI, USA, 2001; Volume 2, pp. 1318–1321. 22. Analog Devices. LF-2.7 GHz RF/IF Gane and Phase Detector (AD8302 Datasheet); Analog Devices: Norwood, MA, USA, 2008. 23. Engen, G.F. The Six-Port Reflectometer: An Alternative Network Analyzer. IEEE Trans. Microw. Theory Tech. 1977, 25, 1075–1080. [CrossRef] 24. Engen, F. An Improved Circuit for Implementing the Six-Port Technique of Microwave Measurements. IEEE Trans. Microw. Theory Tech. 1977, 25, 1080–1083. [CrossRef] 25. Wang, M.; Xue, B.; Wang, W.; Yang, J. The design of multi-user indoor UWB localization system. In Proceedings of the 2nd International Conference on Frontiers of Sensors Technologies (ICFST), Shenzhen, China, 14–16 April 2017; pp. 322–326. 26. Ghannouchi, F.M.; Mohammadi, A. The Six-Port Technique with Microwave and Wireless Application; Artech House: Norwood, MA, USA, 2009. 27. Petrocchi, A.; Raffo, A.; Avolio, G.; Lukasik, K.; Resca, D.; Vannini, G.; Lamey, S.; Caytan, O.; Agneessens, S.; Rogier, H.; et al. An ultra-wideband sensing board for IoT. In Proceedings of the 2017 13th International Conference on Advanced Technologies, Systems and Services in Telecommunications (TELSIKS), Nis, Serbia, 18–20 October 2017; pp. 174–177. 28. Available online: https://www.keysight.com/en/pc-1297113/advanced-design-system-ads?cc=US&lc=eng (accessed on 6 October 2019). 29. Michler, F.; Lindner, S.; Lurz, F.; Erhardt, S.; Weigel, R.; Koelpin, A. Zero-Bias Schottky Power Detector Design for Six-Port based Radar Systems. In Proceedings of the2017 47th European Microwave Conference (EuMC), Nuremberg, Germany, 10–12 October 2017; pp. 228–231. 30. Fan, X.; Pei, X.; Xiong, X. Zero bias Schottky diodes use in high performance detection circuits. In Proceedings of the 2011 International Conference on Electronics and Optoelectronics, Dalian, China, 29–31 July 2011; pp. V3-27–V3-30. 31. Agilent Technologies. The Zero Bias Schottky Diode Detector at Temperature Extremes—Problems and Solutions (Application Note 1090); Agilent Technologies: Santa Clara, CA, USA, 1999. 32. Ikalainen, P.K. An RLC Matching and Application in 1-20 GHz Monolithic Amplifier. In Proceedings of the IEEE MTT-S International Microwave Symposium Digest, Long Beach, CA, USA, 13–15 June 1989; Volume 3, pp. 1115–1118. 33. Ghannouchi, F.M.; Bosisio, R.G. An Alternative Explicit Six-Port Matrix Calibration Formalism Using Five Standards. IEEE Trans. Microw. Theory Tech. 1988, 36, 494–498. [CrossRef] 34. Lemey, S.; Castel, T.; Van Torre, P.; Vervust, T.; Vanfleteren, J.; Demeester, P.; Ginste, D.V.; Rogier, H. Threefold Rotationally Symmetric SIW Antenna Array for Ultra-Short-Range MIMO Communication. IEEE Trans. Antennas Propag. 2016, 64, 1689–1699. [CrossRef] Electronics 2019, 8, 1191 16 of 16

35. Madonna, G.; Ferrero, A.; Pirola, M. Design of a Broadband Multiprobe Reﬂectometer. IEEE Trans. Instrum. Meas. 1999, 48, 622–625. [CrossRef] 36. Petrocchi, A.; Raﬀo, A.; Bosi, G.; Vannini, G.; Kapusuz, K.Y.; Lamey, S.; Rogier, H.; Avolio, G.; Schreurs, D. An ultra-wideband setup to monitor antenna-impedance variations in low-cost IoT transmitters. In Proceedings of the 2018 International Workshop on Integrated Nonlinear Microwave and Millimetre-Wave Circuits (INMMIC), Brive La Gaillarde, France, 5–6 July 2018; pp. 1–3.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).